What is a "Land of Opportunity?"

I just heard Clive Crook on NPR describing new research that shows that by one measure, the United States falls behind Finland, Sweden and France (yes, France) as a land of opportunity. The measure? How well a parent's income predicts his or her offspring's income. Presumably, if we all started out even, if our parents didn't matter, there would be no correlation in income across generations.

But why do kids of the affluent do well in the US? Is it a Paris Hilton effect? Perhaps. If this were the whole story, then America's claim as a land of opportunity is on shaky groud. My suspicion is that things are more complicated.

I happen to live in one of the most affluent counties in the United States--Montgomery County, Maryland. The reason I live here, instead of the District of Columbia, is that the schools here are excellent. One of the reasons that the schools are excellent is that the kids in these schools are amazing. I never fail to be impressed by the work ethic of my sixteen-year olds' circle of classmates. These kids, whose parents are by and large pretty well off, will be by and large well off themselves. Their affluence will not come from clipping bond coupons with golden scissors, but rather from the fact that they will be valuable in the labor force. And as the world globalizes, people with rare skills will become more valuable, as their abilities affect more people.

Paul Krugman suggests that it is policy, rather than productivity, that has led the income distribution to become more skewed, and has reduced social mobility. One wag supported this notion by pointing out that Paul Krugman makes far more money than James Tobin ever did, even though Tobin was a greater economist.

Tobin was the better economist, but Krugman is more productive, because his ideas reach a much broader group of people--technology had enhanced Krugman's productivity and therefore his salary. People with high levels of education and unusual gifts are more valuable than they have ever been. Well educated parents have a particular incentive to make sure that their children gain both the attitudes and credentials necessary for success.

But in the end, this is unfair to those whose parents are at the bottom of the income ladder. Among other things, they cannot afford to live in those parts of metropolitan areas where the schools are excellent. Let me share something I wrote for the Congress on New Urbanism:

Urban Location and Schools

Richard K. Green

The George Washington University

Neighborhoods within cities are heterogeneous; municipalities within metropolitan area are heterogeneous too. The racial and income composition of individual neighborhoods in diverse cities is rarely similar to the composition of the cities in which they lie. For example, the average poverty rate in Washington, D.C. and its proximate counties (Arlington, Fairfax and Alexandria City, Virginia and Montgomery and Prince George’s County, Maryland) was 8.2 percent in 2000; but the rate ranged from 4.5 percent in Fairfax County to 20 percent in the District. Within the District itself, census tract level poverty rates ranged from around one percent to over 70.

The reasons for this variation are myriad. Perhaps the most important reasons for this variation have nothing directly to do with schools: the availability of affordable housing and the cost of transportation. William Alonso, in a classic paper, uses a simple theoretical model that predicts that in a world of high fixed transportation costs, the poor will settle in a city’s center. The reason: by living close to jobs, neighbors and services, the poor avoid transportation costs. In exchange, they consume little in the way of housing, and live very densely. While this leads to housing being relatively inexpensive, the cost they spend per unit of housing services they receive is high. That is, while total rent in the middle of Philadelphia is relatively inexpensive, the houses people are renting may be in disrepair—all of the rent is going to pay for location.

But it is also true that low income people may be constrained to live in urban centers because of a phenomenon William Fischel termed fiscal zoning. Local officials have an incentive to please the median voter in their communities by maximizing services relative to taxes. One tool for doing this is zoning that assures that the least expensive houses are very expensive indeed. Some communities ban apartments; others have large minimum lot size requirements. This combination creates a large per capita tax base and leads to high quality services being capitalized into property values. But it also locks low income people out.

A crude, but easy method, for demonstrating how settlement patterns have played out is to look at the household shares of single women with children and married couples with children in some cities and immediately adjacent suburbs.

Central City County Unmarried Female with Children Share Married Couple with Children Share Suburban County Unmarried Female with Children Share Married Couple with Children Share
Orleans 14.0 16.5 Jefferson 7.2 27.7
Suffolk 9.1 13.6 Middlesex 4.4 24.5
San Francisco 4.3 13.8 San Mateo 4.1 24.5
Philadelphia 9.8 16.9 Montgomery 3.1 28.0
St. Louis 12.0 12.8 St. Louis County 5.3 26.8
Denver 6.7 15.6 Jefferson 5.4 30.1

The poverty rate among married couples with children is low—around 6 percent, or about half the national rate. But the poverty rate among unmarried with children is nearly 35 percent. So the combination of urban land prices, transportation costs and fiscal zoning has led to income heterogeneity in settlement patterns.

While schools may not in themselves be the primary cause of these settlement patterns, the implications of these patterns for schools are important. The ability of low income people to choose to live in areas with good schools is limited by economic considerations apart from school, or by zoning. Conversely, families of means have the ability to choose economically feasible neighborhoods and schools jointly. Let us be clear—school choice is already available to people with money. This lesson came home to me when my family moved to metropolitan Washington; my wife and I were in the fortunate position of being able to choose the public schools our children would attend. The choice set broadened further when my children decided to attend a magnet high school outside of their local school district.

It is unfair that geography allows only some families to choose where their children go to school. Among other things, it means that almost all students at some schools come from families whose choices are limited. A large number of studies (including one by conference participant Thomas Nechyba and his Duke colleague Jacob Vignor) have demonstrated the importance of peer effects—who one goes to school with matters. Of particular interests are findings that students who go to school with peers who have a variety of abilities may perform better, controlling for average level of ability, than those who go to homogeneous schools. Economic and zoning conditions force some families to send their children to schools with students whose performance is both homogeneous and low.

An appropriate policy response to this is to decouple school choice from geography. School vouchers are one method for doing this; charter schools are another; magnet schools are a third—all are worth trying. The Experimental Housing Allowance Program demonstrated that housing vouchers were an efficient and effective method for providing low-income people with housing. The current experiments with choice and charter schools could well prove to me even more important.